Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 1 Gusau Journal of Accounting and Finance (GUJAF) Vol. 2 Issue 3, April, 2021 ISSN: 2756-665X A Publication of Department of Accounting and Finance, Faculty of Management and Social Sciences, Federal University Gusau, Zamfara State –Nigeria Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 2 FAIR VALUE MEASUREMENT HIERARCHY AND COSMETIC ACCOUNTING IN THE NIGERIAN DEPOSIT MONEY BANKS Tesleem Olayinka Adeyemi Department of Accounting Ahmadu Bello University, Zaria-Nigeria. +2348038079555, adeteshplc@yahoo.com, aotesleem@abu.edu.ng Hamisu Suleiman Kargi PhD Department of Accounting Ahmadu Bello University, Zaria – Nigeria. Muhammad Shehu Tijjani PhD Department of Accounting Ahmadu Bello University, Zaria – Nigeria. M. Zubair PhD Department of Business Administration Ahmadu Bello University, Zaria – Nigeria. Abstract The study investigated the effect of fair value financial instruments measurements hierarchy disclosures on cosmetic accounting practices in the Nigerian DMBs, given the possibility of managerial discretion to manipulate the disclosure of fair value measurement, particularly level three fair value hierarchy which is based on adjusted unobservable inputs. The study used a sample of fourteen DMBs in Nigeria that have published their audited annual financial report between 2012 and 2018. The data were subjected to a multiple regression analysis to explore possible effects of fair value measurements on cosmetic accounting. The results revealed that fair value measurements hierarchy significantly reduced the tendency of Nigerian DMBs to manipulate earnings. Specifically, level one and level two fair value measurements which are respectively based on unadjusted and adjusted observable market information were found to be negatively and significantly influencing the level of cosmetic accounting practice among DMBs in Nigeria. On the other hand, result reveals level three fair value measurement is positively and significantly influencing the cosmetic accounting practices. By implication, the findings corroborate the arguments that fair value hierarchy level one and level two will protect accounts from earnings manipulation while level three measurements will enhance unethical accounting practice because the estimate is largely based on adjusted unobservable market information. Therefore, the study recommends the need for regulatory authorities to create an active market for financial instruments in order to fully achieve the fundamental objective of fair value. Also there must be effective supervisory and regulatory framework to limit the uncertainty and ambiguities around the application of level 3 fair value hierarchy. Keywords: Cosmetic accounting, Fair value measurement hierarchy, IFRS, Loan loss provision, Nigerian DMBs . 1. Introduction Managerial opportunism to manipulate accounting information has continue to generate serious concern in corporate finance literature given the continue collapse of companies around the world as a result of unethical accounting practice by those saddle with the responsibility of managing affairs of companies. Cosmetic accounting which can be used interchangeably with earnings managements, creative accounting, hocus pocus accounting, financial engineering, and earnings smoothing, is a flexible accounting instrument use by managers to reduce variability of earnings. Cosmetic accounting or earnings management has always been a subject of hot debate in accounting research and it is widely perceived that under the new IFRS standards, the mailto:adeteshplc@yahoo.com Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 3 introduction of fair value may give room for more earnings manipulations because of the certain discretion allowed under the standards. The world over has experienced a paradigm shift in the determination of income and other elements of financial statements following the adoption of IFRS standards. IFRS comes with a lot of changes in way and manner the information contained in the company’s financial statement are reported. For instance, the introduction of fair value principle, which is regarded as the most important implication of IFRS, motivates more debate on the adoption of the standards. More clearly, IFRS required the usage of fair value contrary to the book value as used by Nigerian GAAP. It is believed that fair value provides up-to-date information about assets as it reflects their real value, hence more relevance for economic decision making (Bello, Abubakar & Adeyemi, 2016). Undoubtedly fair value accounting proved to be more relevant for economic decisions, however, the reliability of its measurement has always been a source of concern and it has been argued in the literature that fair value accounting lacks reliability and consequently can significantly reduce its value relevance (Bosch, 2012). In order to address the contending issues of reliability of fair value measurement, IASB came up with IFRS 13 which sets out a single framework for measuring fair value and specifies the required disclosure about fair value measurement. Consequently, entities applying IFRS now must have to disclose financial instruments measured at fair value based on a three-level hierarchy. According to IFRS 13, companies are expected to disclose the inputs used in measuring the fair value of financial instruments. In order to achieve this, the standard defines a three level measurement hierarchy. Preference is given to unadjusted observable (quoted prices in active markets) level one hierarchy. Level two involves some adjustments on observable inputs from quoted prices of comparable items in active markets, identical items in inactive markets or other market-related information. Level three, on the other hand, involves the use of unobservable (firm generated) inputs in fair value measurements. Given that most of the financial instruments are measured at fair value and the fact that these instruments represent significant part of bank’s financial statements, the impact of fair value accounting especially in financial sectors such as deposit money banks has been a subject of heated debate among academic researchers, investment analysts and policy makers. Empirical studies have shown that banks have incentives to meet regulatory capital requirements and earnings targets, and to reduce taxes. The objectives can be achieved by managing accruals such as loan loss provisions, loan charge-offs, security gains and losses or adjusting investment strategies (Beatty, Ke, and Petroni, 2002). Fundamentally, one of the major concerns about the conversion to IFRS is the issue of fair value accounting which is regarded as the most important implication of IFRS adoption and there is empirical evidence which suggests the possibility of managerial discretion to manipulate the disclosure of fair value measurement. For example, Song, Thomas & Yi (2010), argue that the disclosure of fair values based on less transparent inputs (Level 3 fair values) is less value relevant than the disclosure of fair values based on more transparent inputs (Level 1 and Level 2 fair values). Similarly Hsu and Lin (2016) provide documentary evidence that firms with more Level 3 assets and liabilities (regarded as less value relevant fair value measurement) are more likely to manipulate reported earnings. The discretion inherent in fair value measurement especially in a situation where there is no observable market information provides managers incentive to manipulate the disclosure of fair value measurement. Furthermore, the mark-to- Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 4 model fair value estimates give managers more flexibility to engage in high level of earnings manipulations because they are not based on reliable market information. Thus the very nature of fair value estimate gives managers a great deal of discretion in determining the earnings in any given period. It is expected that most of the companies in developing countries, including Nigeria, will more often, estimate their fair value of financial instruments using adjusted observable or unobservable firm specific generated input (i.e mark-to-model valuation approach) than companies in developed market due to lack of observable market information. Thus, the absence of active markets has led to situation where valuation models are applied which increase the possibility of inherent measurement error in the estimates or management induced error, and this creates opportunity for managers to manipulate with estimation values and consequently results in lower quality of reported earnings. Again, the frequent amendments to fair value standards have made their application very complicated especially for developing countries like Nigeria. In particular, the absence of active markets for financial instruments coupled with weak regulatory environment and fair value assessment gap had actually made it extremely hard for auditors and accountants to do their job and control the fair value measurements (PwC 2015). This argument is aptly corroborated by Benston (2008) who pointed out that fair values other than those taken from quoted prices (level 1) could be readily manipulated by opportunistic and overzealous managers, would be costly to make, and very difficult for auditors to detect and challenge. Furthermore, high cost of fair value estimation may also pose a serious challenge because the estimation requires huge resources such as knowledge in valuation techniques, and special training to enable the auditors and accountants to estimate and audit the financial instruments fairly and objectively. Several studies, particularly in the developed markets, have examined the impact of fair value accounting disclosure on reported earnings with mixed documented evidence. For example, Goh, Li, Ng and Yong, (2015); Ehalaiye (2014); Fiechter (2011); Song et al (2010); Blankespoor, Linsmeier, Petroni, and Shakespeare (2010); Hanselman (2009); Muller, Riedl, and Sellhorn, (2008); Barth (1994); among others, provide empirical evidence that fair value estimations improve the transparency in the financial statements, enhances earnings quality and reduces the level of earnings management. On the other hand, studies such as Alaryan, Haija and Alrabei (2014); Xiaolu (2013); Bratten, Causholli and Myers (2012); Barth, Biscarri, Kasznik and Espinosa (2012); Laux and Leuz (2010); Allen and Carletti (2008); Power (2008); Benston (2006) argued that fair value accounting rather complicates the financial reporting process and makes managerial fraud very difficult to detect thereby increasing the tendency of earnings manipulation by managers. Thus, the fundamental question that is yet to be resolved in the literature is the extent to which fair value measurements hierarchy influences cosmetic accounting practices particular in developing markets with semi-efficient and illiquid market for financial instruments such as Nigeria. Despite the fact that empirical researches concerning the impact of fair value accounting had gained momentum and international relevance especially in the developed economies, however, to the best of our knowledge the effect of fair value financial instruments measurements hierarchy on cosmetic accounting practice has not been tested in developing countries such as Nigeria thus it is not known with certainty whether the theoretical postulations that fair value measurement could be used by managers to manage earnings holds water. This study, therefore, Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 5 attempts to fill this gap by providing empirical evidence on the effect of fair value financial instruments measurements hierarchy on cosmetic accounting practice of Nigerian deposit money banks. The main objective of this study is to examine the effect of fair value financial instruments measurement hierarchy on cosmetic accounting practice of listed Deposit Money Banks in Nigeria. The specific objectives of the study are to: ascertain the effect of fair value measurement hierarchy level one, level two and level three financial assets on cosmetic accounting practice of listed DMBs in Nigeria. Hence, on the bases of the objective of the study, the hypothesis tested is stated as follows: Ho1 Fair value measurement hierarchy level one financial assets has no significant effect on cosmetic accounting practice of listed deposit money banks in Nigeria. Ho2 Fair value measurement hierarchy level two financial assets have no significant effect on cosmetic accounting practice of listed deposit money banks in Nigeria. Ho3 Fair value measurement hierarchy level three financial assets have no significant effect on cosmetic accounting practice of listed deposit money banks in Nigeria. As all the listed firms in Nigeria are mandated to comply with IFRS starting from 1st January, 2012, the study covers 2012-2018. The choice of 2012 to 2018 is based on the ground that the introduction of fair value principle is regarded as the most important implication of IFRS adoption. The study focuses on the listed deposit money banks in Nigeria. This is because significant part of banks’ financial statements consists of financial instruments which are required to be measured at fair value. This study contributes to the ongoing debate concerning the value relevance of fair value measurements hierarchy as required by IFRS 13. In addition, the study contributes to the growing literature on fair value accounting and provides useful information to investors and financial analysts on the implication of new fair value disclosure requirements. More specifically, regulators such as Financial Reporting Council of Nigeria (FRCN), Security and Exchange Commission (SEC), Central Bank of Nigeria (CBN), as well as investors and analysts would find this study of particular interest as it will provide empirical evidence on the usefulness of fair value based accounting estimates. The remaining part of this paper is divided as follows: the review of empirical literature, theoretical framework, methodology, model specification, results and discussion, conclusion and recommendations, and a list references. 2. Literature Review & Theoretical Framework The term cosmetic accounting (CSA) can be referred to as earnings management, window dressing accounting; creative accounting, financial engineering, accounting hocus-pocus. However, the preferred term in most of the literatures is earnings management. According to Copeland (1968) “earnings management involves the repetitive selection of accounting measurement or reporting rules in a particular pattern, the effect of which is to report a stream of income with a smaller variation from trend than would otherwise have appeared. Earnings management is a strategy employed by management of a company to deliberately manipulate the company’s earnings so that the figures match a predetermined target. Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 6 Fair value in Accounting Standards has been a subject of serious concern following the adoption of IFRS standards. Prior to the adoption of IFRS, International Generally Accepted Accounting Principles (IGAAP) defined Fair value of an asset as the amount for which that asset could be exchanged between knowledgeable, willing parties in an arm's length transaction. The adoption of IFRS has brought about some modification in the definition of fair value in a logical and comprehensive manner. IASB conceptualizes fair value as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date (i.e. it is an exit price). Thus, IASB definition clearly emphasizes that fair value is essentially a market-based measurement, not an entity-specific measurement. Moreover, the new definition of fair value explicitly focuses more on an orderly transaction and the measurement date. The use of fair value measurements to manipulate reported earnings has been widely discussed in the literature particularly in the developed countries. Several attempts have been made to provide empirical evidence regarding manager’s incentive to manipulate the disclosure of fair value measurement in order to manipulate the reported earnings. Numerous researchers provide empirical evidence that fair value estimations improve the transparency in the financial statements, enhances earnings quality and reduced the level of earnings management. For instance, Ehalaiye (2014); observed strong predictive relationship between fair value accounting and banks performance taking into the consideration the samples of US banks. Also, Fiechter (2011) examined the fair value option on earnings volatility in Europe. The study provides empirical evidence that the effects of fair value option are significantly relevant and its application has improved reporting of the true economic consequences of financial transactions as well as improved timeliness of financial statements. Song et al (2010) examined the value relevance of FAS 157 fair value hierarchy information and the impact of corporate governance mechanisms using quarterly data of US quoted firms in the year 2008. The study revealed that fair value disclosures are overall value relevant and improved the quality of reported earnings. However, the value relevance of Level 1 and Level 2 fair values was found to be greater than the value relevance of Level 3 fair values. In addition, the study found that the value relevance of fair values (especially Level 3 fair values) is greater for firms with strong corporate governance. On the other hand, several studies provide evidence that fair value accounting rather complicates the financial reporting process and makes managerial fraud difficult to detect thereby increasing the tendency of earnings manipulation by managers. Alaryan et al. (2014) examined the relationship between fair value accounting and the presence of earnings manipulation using annual report of ten year period (1997-2006) split into five year before and after the adoption of the standards. The results indicate that the number of firm that manipulated their financial statement information had increased after the application of fair value accounting. Bratten et al. (2012) examine the association between the magnitude of fair value reporting and bank earnings management through discretionary loan loss provisions and discretionary security gains and losses. The study documents that banks whose auditors are industry specialists are less likely to manage earnings. Similarly, Barth et al. (2012) using comprehensive data from US commercial banks and bank holding companies, provide evidence that fair value accounting increases the tendency of earnings management by managers, in particular the study revealed that fair value gains in AFS assets have consistently been used for earnings and capital management and that the holdings of AFS assets are related to the intensity earnings Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 7 manipulation by management. In clear term, the more available-for-sales assets a bank holds, the greater “opportunity” that a bank uses realized gains and losses to smooth earnings. In addition, the results show that the earnings management behavior is present both in listed and non listed banks, implying that the motivations go beyond the incentives provided by capital markets. Other studies such as Xiaolu Xu (2013); Laux and Leuz (2010); Allen and Carletti (2008); Power (2008); Benston (2006) provide documentary evidence regarding manager’s incentive to manipulate the disclosure of fair value measurements to achieve a predetermined objective or to beat analyst forecast target. The study adopts agency theory to underpin the study. Agency theory describes the relationship which exists between the principal (shareholders) and the agents (management) whereby the management direct the affairs of company on behalf of the shareholders. The theory views directors as the agent of the shareholders and as such they are expected to act in the best interest of the shareholders. Sometimes, the directors are motivated to act in their own best interest and this creates a conflict between the interest of shareholders and that of the directors. These conflicts usually arise when directors and shareholders have different interests and there exists information asymmetry (i.e. the directors having more information). Information asymmetry may result between the contracting parties as managers may be in possession of superior information about the present and expected future performance of the entity than the owners. This may incentivize managers to portray a favorable picture of the entity for their personal benefit. Therefore, agency theory raises a fundamental problem in organizations self-interest behaviuor and thus stresses the separation of ownership (principal) and directors (agent) in an organization. The shareholders delegated authority of the management of the company to the directors, therefore, it is expected that the directors act in the best interest of the shareholders. However, it is believed that directors may sometimes take decisions which may conflict with the interest of the shareholders. Arguably, managers could be tempted to manipulate fair value estimates that promote their interest leading to biases in the information presented in the entity’s financial statement. Documented evidence have shown that when accounting information is highly subjective and managers discretion allowed, intentional biases in the accounting aggregate estimates is very likely (Ehalaiye, 2014). 3. Methods and Techniques This study employed correlation research design. This is concerned with the collection of data for the purpose of describing and analyzing the impact of fair value accounting on cosmetic accounting of quoted deposit money banks in Nigeria. The data for this study were obtained mainly from secondary sources which were extracted from the audited annual reports and accounts of quoted DMBs in Nigeria from 2012 to 2018. The study population consist of all the fifteen Deposit Money Banks listed on the Nigerian Stock Exchange as at 31 st December, 2012 and remained listed up till 2018. Using census approach, all the fourteen listed Deposit Money Banks in Nigeria as at 31st December, 2018 were used for the analysis due to availability of their annual reports and accounts needed for the extraction of the data. In analyzing the data for this study, a panel data multiple regression technique and descriptive statistics was used. By the dependent variable the study used discretionary loan loss provision as a proxy for cosmetic accounting. The independent variables are Fair value level one, fair value level two and fair value level three financial assets measurements hierarchy. Both financial leverage and bank size are included in the model as control variables. To test the hypotheses, panel multiple regression models with an error term is specified in econometric form as shown below: Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 8 CSAit = β0 + β1FVA1it + β2FVA2it + β3FVA3it + β4FLEVit + β5FSIZEit + Cosmetic Accounting Measurement The study employed Chang, Shen and Fang (2008) model of discretionary loan loss provision which was specifically built for financial sector. The residual from this model was used to represent cosmetic accounting. DLLPi /TAt-1 = LLPit/TAt-1 – {α0 1/TAt-1 + α1LCOi/TAt-1 + α2BBALi/TAt-1}……………………(i) Where DLLP = Discretionary loan loss provision LLP = Loan loss provision LCO = Loan Charge-off BBAL = Beginning Balance of loan loss TAt-1 = Lagged Total Assets α0= Constant Measurement of Explanatory Variables Variable Acronym Definition Source Level one fair value assets FVA1 Level one fair value assets divided by the total assets Hsu and Lin (2016) Level two fair value assets FVA2 Level two fair value assets divided by the total assets Hsu and Lin (2016) Level three fair value assets FVA3 Level three fair value assets divided by the total assets Hsu and Lin (2016) Firm Size FSIZE Log of total assets. Financial Leverage FLEV Measured as the ratio of total debts to total assets 4. Results and Discussions 4.1 Statistical Criterion The results in table 1 provide descriptive statistics of variables, where the minimum, maximum mean and standard deviations of the data are fully presented. Table 1: Summary of Descriptive Statistics Variables Observations Mean Std Dev Min Max CSA 98 0.02 0.02 0 0.14 FVA1 98 0.05 0.09 0 0.61 FVA2 98 0.01 0.01 0 0.05 FVA3 98 0.01 0.03 0 0.19 FSIZE 98 8.36 1.17 5.95 9.77 FLEV 98 86.06 4.29 71.72 93.65 Source: STATA Output, 2021 The results in Table 1 above provide some insight into the nature of quoted deposit money banks that reported their financial in line with IFRS 13 financial instrument measurement hierarchy for the period 2012 to 2018. It shows the mean (average), standard deviation (degree of dispersion), the maximum and minimum for each of the variables. It reveals average cosmetic accounting i t  Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 9 practice (CSA) of 2% of total lagged asset of the sample banks with a standard deviation of 0.02. The minimum is 0 while the maximum is 0.143. Average FVA1 is approximately 5%, the standard deviation is 0.09 and range between 0.0 and 61%. FVA2 has a mean of 0.6% and the standard deviation is 0.01 and the minimum and maximum are 0.0% and 5 percent respectively. The mean value of FVA3 is about 0.1%. This means that very low proportion of bank’s financial assets is classified under level three fair value measurement. The minimum is 0.0% while the maximum is 20%. Furthermore, FSIZE has an average of 8.4 revealing that Nigerian banks are large in terms of capital base. The minimum is 5.95 while the maximum is 9.77. It can be seen that the averages of the variables do not differ substantially from their respective standard deviations which means that the data are not skewed and are fit to produce a reliable result. Table 2: Correlation Matrix VARIABLES CSA FVA1 FVA2 FVA3 FSIZE FLEV CSA 1 FVA1 -0.167* 1 FVA2 -0.088 0.002 1 FVA3 0.127 0.113 0.067 FSIZE -0.123 -0.097 -0.416*** 0.133 1 FLEV 0.065 0.014 0.033** 0.054 -0.287 1 ***Correlation is significant at the 0.01 level (2-tailed) **Correlation is significant at the 0.05 level (2-tailed) Source: Source: STATA Output, 2021 Correlation matrix shows the relationship between explanatory variables and explained variable and also the relationship among the individual variables themselves. The results from Pearson correlation indicates a negative and significant association between CSA and FVA1 and negative and insignificant association between CSA and FVA2. The result also indicates a positive and insignificant association between CSA and FVA3. Amongst the independent variables, the relationship was a very weak one as expected which may not pose any multicollinearity problem. Gujarati (2004) opines that correlation above 0.8 between variables is a concern as it indicates excessive correlation. The tolerance values and the variance inflation factor are an important measure of multicolinearity between the independent variables in a study. The results indicate that variance inflation factor were consistently smaller than 10 indicating absence of multicolinearity problem. This shows the suitability of the study model with all the explanatory variables used in the study. Further, the tolerance values were consistently smaller than 1.00, therefore, substantiating the fact that there is complete absence of multiconearity between the explained and the explanatory variables. See appendices for stata output. Table 3: Regression Result Variables Coefficient Std error t-value p-value FVA1 -0.0571 0.0547 -3.69 0.000 FVA2 -0.5051 0.1979 -2.55 0.012 FVA3 0.1692 0.0534 3.17 0.002 FSIZE -0.0056 0.0021 -2.64 0.010 FLEV -0.0001 0.0007 -0.09 0.927 Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 10 Constant 0.0797 0.0669 0.24 0.237 R 2 0.11 F-Stat. 5.73 F-Sig. 0.000 Source: Source: STATA Output, 2021 From the result in Table 3, it can be observed that the R-squared which is the multiple coefficient of determination is 0.11. This implies that about 11% of the total variation in CSA of listed deposit money banks in Nigeria is jointly explained by all the explanatory variables included in the model of the study, while the remaining 89% was caused by factors not captured in the model. The F-statistic is 5.73 which is significance at one percent, shows that the model of the study is fit and all the explanatory variables were properly selected, combined and used. From the results, level one fair value measurement was found to be negatively and significantly influence on cosmetic accounting practices in DMBs with a coefficient of -0.06 and t-value of - 3.69 which is significant at 1%. This suggests that an approximately 1% increases in level one fair value measurement results in approximately 7kobo decreases in cosmetic accounting practice. Furthermore, the negative association between cosmetic accounting and level one fair value measurement implies that as more financial assets of banks are measured at fair value using the observable market input the less the possibility of earnings manipulation. The result is not surprising because level one fair value measurement is adjudged to be more transparent and based on the observable market information; consequently the tendency of earnings manipulation through the discretionary measurement would be significantly reduced. This finding provides reasonable and documentary evidence to reject the hypothesis that level one fair value measurement does not significantly affect cosmetic accounting practices of quoted DMBs in Nigeria. The finding is consistent with prior literature on fair value measurement such as Goh et al (2015); Ehalaiye (2014); Fiechter (2011); Song et al (2010) among others, who provide empirical evidence that fair value estimations improves transparency in the financial statements, enhances earnings quality and reduced the level of earnings management. In particular, level one fair value measurements reduce the possibility of earnings manipulation and enhance the quality of accounting information. On the other hand, it contradicts the findings of Bratten et al. (2012); Barth et al (2012); Xiaolu (2013); Laux and Leuz (2010); who provide documentary evidence regarding manager’s incentive to manipulate the disclosure of fair value measurements to achieve a predetermined objective or to beat analyst forecast target. The results from Table 3 also revealed a negative and significant association between level two fair value measurements and cosmetic accounting practice in Nigerian DMBs with a coefficient of -0.5 and a t-value of -2.6 which is significant at 1% level implying that N1 increases in level two fair value measurement results in a decrease of cosmetic accounting practice by about 50kobo. This suggests that the use of level two fair value estimates which is based on the adjusted observable input lowers the level of unethical accounting practice of listed DMBs in Nigeria. This finding further provides empirical evidence that more transparent inputs of level one and level two fair value measurements deter manager’s manipulation. The finding of this study provides reasonable and valid evidence to reject the hypothesis that level two fair value measurement does not significantly affect cosmetic practices of quoted DMBs in Nigeria. Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 11 However and expectedly, the result further reveals a positive and significant relationship between level three fair value measurements and cosmetic accounting practice having a coefficient of 0.17 and a t-value of 3.17. This implies that level three fair value measurements significantly influence cosmetic accounting practices in Nigerian DMBs. This is in line with expectation that more level three fair value measurements will lead to a rise in cosmetic accounting practices because it is based on unobservable firm-generated inputs which provides manager greater flexibility in managing the valuation of level three financial instruments. Further, the result is not surprising because of the current developing nature of Nigerian capital market and the fact that there is illiquid market for financial instruments in Nigeria. Again, the finding is in line with our prior expectation and therefore, provides a reasonable ground to reject the hypotheses that level three fair value measurements has no significant effect on earnings cosmetic accounting practice of listed DMBs in Nigeria. As for control variables, we observe that bank size inversely and significantly influence the level of cosmetic accounting practice. This means that the size of the bank can influence the management policy of the company. This is in line with the view that because large banks have more resources than their smaller counterparts as such they are more likely to avoid the use of fair value measurements to manipulate reported earnings. 5. Conclusion and Recommendations In this study, an attempt was made to examine the impact of fair value measurements hierarchy on cosmetic accounting practice in Nigeria. The empirical research of this study is based on the sample of 14 DMBs for the financial years 2012 to 2018. Using a panel data multiple regression model, the study provides strong evidence that level one and level two fair value measurements inversely and significantly influencing cosmetic accounting practices in the Nigerian DMBs. This implies that more transparent inputs of level one and level two fair value measurements deter managers’ manipulation. The result also reveals that level three fair value measurements which is based on unobservable input negatively and significantly influencing the level of cosmetic accounting practice in Nigerian DMBs. Therefore, the study recommends the need for regulatory authorities to create an active market for financial instruments in order to fully achieve the fundamental objective of fair value. Also there is need for effective supervisory and regulatory framework to limit the uncertainty and ambiguities around the application of level 3 fair value hierarchy. References Alaryan, L. A., Haija, A. A. A., &Alrabei A. M., (2014). The relationship between fair value accounting and the presence of manipulation in financial repoting. International journal of accounting and financial reporting. Vol. 4(1). Allen, F. and E. Carletti (2008) “Mark-to-Market Accounting and Liquidity Pricing,” Journal of Accounting and Economics 45, 358–378. Barth, M. E., J. Gomez-Biscarri, R. Kasznik and G. Lopez-Espinosa. (2012). Fair value accounting, earnings management and the use of available-for-sale instruments by bank managers. Working paper, Stanford University. Barth, M., Clinch, G., (1998). Revalued financial, tangible, and intangible assets: associations with share prices and non-marketbased value estimates. J. Acc. Res. 36 (Supplement), 199– 233. Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 12 Barth, M. E. (1994). Fair value accounting: Evidence from investment securities and the market valuation of banks. The Accounting Review, 69(1), 1-25. Beatty, A. L., B. Ke, and K. R. Petroni. (2002). Earnings management to avoid earnings declines across publicly and privately held banks. The Accounting Review 77 (3): 547-570. Benston, G. J. 2008. The shortcomings of fair-value accounting described in SFAS 157. Journal of Accounting and Public Policy 27 (2): 101-114. Benston, G. J. (2006). Fair-value accounting: A cautionary tale from Enron. Journal of Accounting & Public Policy, 25(1), 465-484. Blankespoor, E., Linsmeier, T.J., Petroni, K.R. and C. Shakespeare (2010) “Fair Value Accounting for Financial Instruments: does it improve the Association between Bank Leverage and Credit Risk?” Available at SSRN: http://ssrn.com/abstract=1565653. Bratten, B., M. Causholli and L. Myers. (2012). The effect of fair value accounting and auditor specialization on earnings management: evidence from the banking industry. Working Paper, University of Kentucky. Chang, R.D, Shen, W. H. & Fang, C. J. (2008). Discretionary loan loss provisions and earnings management for the banking industry. International Business & Economics Research Journal, 7(3): 9-20. Ehalaiye, O. O. (2014). An evaluation of the predictive value of bank fair values, A Thesis submitted to the Victoria University of Wellington USA in fulfilment of the requirements for the degree of Doctor of Philosophy Fiechter, P. (2011). The Effects of the fair value option under IAS 39 on the volatility of bank earnings. Journal of International Accounting Research, 10(1), 85–108. Goh, B. W., Li, D., Ng, J., & Owyong, K. K., (2015). Market pricing of banks’ fair value assets reported under SFAS 157 since the 2008 financial crisis. Singapore Management University. Gujarati, N. D. (2004) Basic economericts. 4 th Edition, McGraw-Hills, USA. Hanselman, O. (2009). Full fair value accounting: Its time has come. Journal of Performance Management, 1, 1-18. Hsu P.H., and Lin Y., (2016). Fair value accounting, earnings management. Eurasian. Journal of Business and Management, 4(2). Laux, C., &Leuz, C. (2010).Did fair-value accounting contribute to the financial crisis. Journal of Economic Perspectives, 24(1), 93–118. Muller, K. A., Riedl, E. J., &Sellhorn, T. (2008).Consequences of voluntary and mandatory fair value accounting: Evidence surrounding IFRS adoption in the EU Real Estate industry. Working Paper, Harvard Business School. Retrieved from: http://www.hbs.edu/research/pdf/09-033.pdf Power, L. (2008).Fair value – friend or foe. Retrieved from: http://www.pwc.com/extweb/pwcpublications.nsf/dfeb71994ed9bd4d802571490030862f /3ea3bd0a293c335e8025749f0032fd8f/$file/fair_value_friend_or_foe_august_2008.pdf PWC (2015) Fair Value Measurements, global edition cfodirect Net work, international pwc sites Song, C.J., Thomas, W.B., Yi, H., (2010). Value relevance of FAS No. 157 fair value hierarchy information and the impact of corporate governance mechanisms. Acc. Rev. 85 (4), 1375–1410. Xiaolu Xu (2013).Fair value measurements and earnings management: evidence from the banking industry. Accounting Dissertations - Whitman School of Management. http://ssrn.com/abstract=1565653 http://www.hbs.edu/research/pdf/09-033.pdf http://www.pwc.com/extweb/pwcpublications.nsf/dfeb71994ed9bd4d802571490030862f Gusau Journal of Accounting and Finance, Vol. 2, Issue 3, April, 2021 13